Overfeeding Polyunsaturated and Saturated Fat Causes Distinct Effects on Liver and Visceral Fat Accumulation in Humans

Fredrik Rosqvist; David Iggman; Joel Kullberg; Jonathan Cedernaes; Hans-Erik Johansson; Anders Larsson; Lars Johansson; Håkan Ahlström; Peter Arner; Ingrid Dahlman; and Ulf Risérus


Diabetes. 2014;63(7):2356-2368. 

In This Article


Despite comparable weight gain after 49 days, this double-blind trial showed that overeating energy from PUFAs prevented deposition of liver fat and visceral and total fat compared with SFAs. Excess energy from SFAs caused an increase of liver fat compared with PUFAs. Further, the inhibitory effect of PUFAs on ectopic fat was accompanied by an augmented increase in lean tissue and less total body fat deposition compared with SFAs. Thus, the type of fat in the diet seems to be a novel and important determinant of liver fat accumulation, fat distribution, and body composition during moderate weight gain. We also observed fatty acid–dependent differences in adipose tissue gene expression. The significant decrease in pancreatic fat in both groups during weight gain was an unexpected finding that needs confirmation due to the low amounts of pancreatic fat in this lean population.

Cross-sectional studies have shown that patients with higher SFA and lower PUFA intake have increased liver fat content,[13,15,25] which is also in accordance with lower PUFA levels in fatty livers.[14,26] A previous isocaloric trial in abdominally obese subjects indicated that the present associations may be causal, since replacing SFAs from butter with PUFAs from sunflower oil reduced liver fat.[20,22] Thus, together these trials indicate that SFAs (high in 16:0) per se might promote hepatic steatosis, both during isocaloric and hypercaloric conditions. These results also support the current nutritional recommendations in general (i.e., to partly replace SFAs with PUFAs). PUFAs (i.e., linoleic acid) are found in plant-based foods such as nuts, seeds, and nontropical vegetable oils.[27] Increased intake of these foods has in general been associated with cardiometabolic benefits including lowering blood lipids and reduced risk of cardiovascular disease and type 2 diabetes.[27–29] There are, however, no clear reasons to believe that sunflower oil would be more effective in preventing liver fat accumulation than other PUFA-rich oils and fats.

The mechanisms behind the differential effects on liver fat deposition are unknown, but may involve differences in hepatic lipogenesis and/or fatty acid oxidation and storage.[30] In NAFLD patients, increased de novo lipogenesis is a major contributor to liver fat accumulation and steatosis.[31,32] In the current study, a fructose–SFA interaction on liver fat is possible since the muffins contained significant amounts of fructose.[33] Early animal data showed that carbohydrate-induced lipogenesis was inhibited by adding linoleic acid, whereas palmitate had no effect,[34] and SFAs have enhanced steatosis and increased hepatic lipogenesis compared with PUFAs.[20,21] Hepatic activity of the lipogenic enzyme SCD-1 may be elevated in steatosis.[26] Also, SCD-1–deficient mice were protected against hepatic lipogenesis, whereas SCD-1 inhibitors markedly reduced hepatic triglyceride accumulation.[35] In humans, a strong association between the change in liver fat and the change in hepatic SCD-1 index was reported in weight-stable subjects,[22] a finding currently confirmed during hypercaloric conditions.

PUFAs are more readily oxidized than SFAs,[36–38] thereby potentially lowering hepatic exposure to nonesterified fatty acids, a major substrate in triglyceride synthesis. Concentrations of D-3-hydroxybutyrate were, however, if anything, lower with PUFAs than SFAs, thus not supporting a differential effect on hepatic fat oxidation. Animal studies have also indicated that SFAs, compared with PUFAs, lower brown tissue adipose activity and thermogenesis.[16–19,39–45]

The increase in lean tissue was nearly threefold higher during PUFA overeating compared with SFA. Although lean tissue was a secondary outcome, this finding is intriguing since obese persons with reduced lean tissue (sarcopenic obesity) are more insulin-resistant and at higher risk for physical disability.[46,47] A previous supplementation trial in postmenopausal women reported that a daily dose of 8 g PUFA (safflower oil) increased lean tissue and reduced trunk fat.[48] In accordance, rats isocalorically fed with PUFAs (high in linoleic acid) gained more lean tissue and less fat compared with an SFA-rich diet, in line with similar studies.[16,17,49,50] The mechanism behind these observations remains to be determined. The differential increase in lean tissue was consistent when assessed by two different methods (MRI and Bod Pod). This difference was unlikely an artifact due to changes in total-body water content since the results were similar in the three-compartment model. Although supported by animal studies, this finding needs to be replicated in additional human studies.

In the current study, n-6 PUFAs were investigated, but it is possible that n-3 PUFAs have similar effects on body fat accumulation.[50–52] The amount of sunflower oil used in the current study (~40 g per day) corresponds to about three times the customary intake of linoleic acid in the Swedish population. Given that palm oil was used as the SFA source, the wide use of this oil by the food industry may be of concern. In fact, palm oil is one of the most used oils worldwide, suggesting a potential global impact if it promotes adiposity. The health effects of palm oil, however, remain uncertain and should be further investigated. The effects on ectopic fat deposition observed in this study, however, do not seem to be palm oil–specific, but rather SFA- or palmitate-specific since we previously showed similar results during isocaloric conditions using butter as the source of SFAs.[22]

Given the different influence on fat deposition, we expected diet-specific influences on adipose gene expression. Overall, differences in SAT gene expression between diets were modest, which may relate to similar weight gain and little differences in SAT. Although speculative, downregulation of ALDH1A1 by PUFAs might be relevant, as this gene inhibits energy dissipation and promotes fat storage.[53] Interestingly, ALDH1A1-deficient mice are protected from diet-induced liver fat accumulation and insulin resistance.[53] The observed associations between changes in SAT fatty acids and mRNA expression support a direct influence of the fatty acids consumed on adipose tissue gene expression. For example, ALDH1A1 was inversely associated with changes in linoleic acid, but directly associated with the SCD-1 index. As gene expression was measured only in SAT, the gene expression results cannot be directly extrapolated to other depots, such as visceral adipose tissue (VAT) and liver fat. Firm conclusions about the mechanisms of PUFA-induced changes in liver metabolism can therefore not be drawn from the current study. These findings thus need confirmation in VAT and liver, which may not be feasible in humans. However, a recent animal study[54] investigated the effect of overfeeding rats with different types of fat varying in linoleic acid content. Rats fed a diet higher in PUFAs (linoleic acid) showed lower liver fat accumulation together with lower hepatic gene expression of several fatty acid transporters (FATP-2, FATP-5, and CD36) and lipogenic enzymes (fatty acid synthase, acetyl-CoA carboxylase, and SCD-1) compared with rats fed a diet lower in linoleic acid. Hepatic gene expression of carbohydrate-responsive element–binding protein and sterol regulatory element–binding protein-1c were also lower in rats fed a diet higher in linoleic acid. Accordingly, we observed that the estimated SCD-1 activity in plasma cholesterol esters (reflecting hepatic metabolism) was markedly decreased in the PUFA group ( Table 4 ), implying that the mechanisms may be at least partly similar (i.e., decreased hepatic lipogenesis).

Some strengths of this study should be mentioned. This study was double-blinded, which rarely is feasible in dietary interventions that include foods rather than supplements or capsules. Our body composition data are strengthened by consistent findings using two independent methods (MRI and Bod Pod). All subjects completed the trial. Both groups in the current study consumed vegetable oils without any cholesterol, thus excluding any confounding effect of dietary cholesterol[55] that is abundant in SFAs from animal sources. Assessment of fatty acid composition in plasma lipids and adipose tissue suggested high adherence to the interventions in both groups. Accelerometer monitoring suggested no bias due to differences in physical activity between groups. As we compared two common dietary fatty acids (the major PUFA, linoleic acid, and the major SFA, palmitic acid) in the Western diet, the results of this study could be relevant to many populations.

This study also has several potential limitations. Notably, our results may not apply to obese or insulin-resistant individuals who might show a different response to the diets, both with regard to ectopic fat accumulation and glucose metabolism. Also, the current healthy, young, and overall lean individuals had very low liver and visceral fat content at baseline. Thus, the lack of differences in fasting insulin concentrations were not surprising (i.e., the absolute increase of liver fat during SFA treatment was most likely too small to produce significant metabolic differences between the diets in this healthy study group). It should, however, be noted that the study was not designed or powered to examine differences in insulin sensitivity, and we did not measure hepatic or whole-body insulin sensitivity directly, which lowered the ability to detect any possible differences between groups. The data thus need confirmation in older individuals with NAFLD or type 2 diabetes and in other ethnic groups. The short duration of the study may not resemble long-term effects. However, results on liver fat are strongly supported by similar effects reported in weight-stable obese subjects, in which also modest effects on insulin levels and triglycerides were observed.[22] The MRI methods used relied on fixed-spectrum models and thus did not allow full characterization of all lipid resonances of the liver spectra to detect changes in liver lipid saturation. However, results from plethysmography were consistent with MRI results regarding body fat deposition. Finally, it should be noted that sunflower oil contains more vitamin E than palm oil, and vitamin E supplementation has decreased steatosis.[56] However, the present vitamin E levels were most likely too low to have an effect, and there was no correlation between change in liver fat and change in vitamin E intake (data not shown). Furthermore, the effects of PUFAs were not exclusive to liver fat.

In conclusion, overeating different types of fat seems to have different anabolic effects in the body. The fate of SFAs appears to be ectopic and general fat accumulation, whereas PUFAs instead promote lean tissue in healthy subjects. Given a detrimental role of liver fat and visceral fat in diabetes, the potential of early prevention of ectopic fat and hepatic steatosis by replacing some SFAs with PUFAs in the diet should be further investigated.